US20080129335A1

Movatterモバイル変換

Info

Publication number: US20080129335A1
Application number: US11/942,691
Authority: US
Inventors: Brad Hutchings; Herman Schmit; Steven Teig
Original assignee: Individual
Current assignee: Altera Corp
Priority date: 2005-03-15
Filing date: 2007-11-19
Publication date: 2008-06-05
Anticipated expiration: 2025-03-15
Also published as: US20070241784A1; US7932742B2; US7310003B2

Abstract

Some embodiments of the invention provide a configurable integrated circuit (“IC”). The IC includes a first set of circuits and a second set of circuits interspersed among the first set of circuits. Each set of circuits includes at least ten volatile configurable circuits. Several circuits in at least one of the sets are user multiplexers. Each particular user multiplexer has input and output terminals and has a set of select terminals for receiving a set of user-design signals that directs the particular multiplexer to connect a set of the input terminals to a set of the output terminals. The user-design signals are signals generated internally by the IC.

Description

FIELD OF THE INVENTION

The present invention is directed towards configurable integrated circuits with interconnect circuits that have select lines driven by user signals.

BACKGROUND OF THE INVENTION

The use of configurable integrated circuits (“IC's”) has dramatically increased in recent years. One example of a configurable IC is a field programmable gate array (“FPGA”). An FPGA is a field programmable IC that has an internal array of logic circuits (also called logic blocks). These logic circuits are connected together through numerous interconnect circuits (also called interconnects). The logic and interconnect circuits are surrounded by input/output blocks.

Like some other configurable IC's, the logic circuits and interconnect circuits of an FPGA are configurable. A configurable logic circuit can be configured to perform a number of different functions. Such a logic circuit typically receives input and configuration data. From the set of functions that the logic circuit can perform, the configuration data specifies a particular function that the logic circuit has to perform on the input data. A configurable interconnect circuit connects a set of input data to a set of output data. Such a circuit receives configuration data that specify how the interconnect circuit should connect its input data to its output data.

FPGA's have become popular as their configurable logic and interconnect circuits allow the FPGA's to be adaptively configured by system manufacturers for their particular applications. However, existing FPGA's as well as other existing configurable IC's do not provide robust multiplexer functionality. Specifically, the select lines of multiplexers in existing IC's are typically tied to memory cells that store configuration data. Hence, these multiplexers cannot be controlled by signals internally computed by the logic circuits of the IC's. This, in turn, limits the use of multiplexers in configurable IC's to applications that do not need to make the multiplexing choices based on internally computed IC signals.

Therefore, there is a need in the art for configurable IC's that use novel multiplexer circuits that can be controlled by signals internally computed by the IC's. There is also a need in the art for configurable IC's that have novel architectures that use such multiplexer circuits. There is further a need for configurable IC's that have devices that can take advantage of such novel multiplexer circuits.

SUMMARY OF THE INVENTION

BRIEF DESCRIPTION OF THE DRAWINGS

The novel features of the invention are set forth in the appended claims. However, for purpose of explanation, several embodiments of the invention are set forth in the following figures.

FIG. 1 illustrates a block diagram of an interconnect circuit.

FIG. 2 illustrates an example of a multiplexer.

FIG. 3 illustrates an example of a configurable four-to-one multiplexer that receives configuration data from a set of storage elements.

FIG. 4 illustrates a configurable logic circuit.

FIG. 5 illustrates an example of a configurable node array that includes configurable nodes that are arranged in rows and columns.

FIG. 6 illustrates an example of a connection between two circuits.

FIG. 7 illustrates an example of another connection between two circuits.

FIG. 8 illustrates a full UMUX.

FIG. 9 illustrates a hybrid UMUX.

FIG. 10 illustrates that some embodiments place the UMUX's between the circuits of a circuit arrangement according to no particular pattern.

FIG. 11 illustrates that some embodiments place UMUX's between the circuits according to a particular pattern that repeats across the entire circuit arrangement or a portion of the arrangement.

FIG. 12 illustrates an example where a number of UMUX's that are dispersed between the circuits of a circuit arrangement, form an arrangement of their own.

FIG. 13 illustrates an example of arranging UMUX's in two arrays that are dispersed between the circuits of a circuit arrangement.

FIG. 14 illustrates an example of a UMUX that is positioned in an arrangement of circuit elements.

FIG. 15 illustrates a circuit arrangement having logic circuits, interconnect circuits, and UMUX's.

FIG. 16 illustrates an example of a circuit in a configurable IC's circuit arrangement that includes complex logic circuits that are formed by multiple logic and interconnect circuits.

FIG. 17 illustrates an example of “supering” multiple logic circuits that use of UMUX's.

FIG. 18 illustrates one such prior configurable circuit arrangement.

FIG. 19 illustrates a configurable circuit arrangement that includes numerous configurable interconnect circuits, UMUX's, and logic circuits.

FIG. 20 illustrates an example of using UMUX's to write data to a non-configurable memory array in a configurable circuit arrangement.

FIG. 21 illustrates the use of UMUX's to implement circuit arrangements with configurable shift operations.

FIG. 22 illustrates an example of a circuit arrangement that includes numerous logic circuits, interconnect circuits, and HUMUX'S.

FIG. 23 illustrates an example of a circuit arrangement with numerous logic circuits and routing interconnect circuits.

FIG. 24 illustrates a circuit arrangement that has numerous logic circuits and interconnect circuits arranged in numerous rows and columns.

FIG. 25 illustrates a two-input multiplexer that receives variables “b” and “c” on its input lines and variable “a” on its select line.

FIG. 26 illustrates that the multiplexer can perform the AND function by routing abinary 0 to its second input.

FIG. 27 illustrates that routing abinary 1 to the multiplexer's first input, allows the multiplexer to perform the OR function, which is one of the NPN equivalent functions of the AND function.

FIG. 28 illustrates a two-input multiplexer that implements an XOR function.

FIG. 29 illustrates an example of an HUMUX that has a core multiplexer that operates like the multiplexer ofFIGS. 25-27 as it receives the user signal “a” through the select multiplexer

FIG. 30 illustrates another example of an HUMUX that has a core multiplexer that operates like the multiplexer ofFIG. 28 as it receives the user signal “a” through the select multiplexer.

FIG. 31 illustrates a recursive process that a synthesizer can perform to decompose a function.

FIG. 32 illustrates an eight-to-one HUMUX that is formed by four multiplexers.

FIG. 33 illustrates a multiplexer that is used as the multiplexer inFIG. 32.

FIG. 34 illustrates a CPL implementation of a two-tier multiplexer structure for generating the second signal S1 and its complement.

FIG. 35 illustrates an example of the signals CLK, ST0, and ST1.

FIG. 36 illustrates a circuit implementation of the combined multiplexer structure.

FIG. 37 illustrates another implementation of an HUMUX, which can operate based on a configuration-derived select signal S0, a first user signal US0, or a second user signal US1.

FIG. 38 illustrates a circuit implementation of the combined multiplexer structure ofFIG. 37.

FIG. 39 illustrates another alternative circuit structure for one set of multiplexers in a two-tiered multiplexer structure of some embodiments.

FIG. 40 illustrates an architecture that is formed by numerous configurable tiles that are arranged in an arrangement with multiple rows and columns.

FIGS. 41-45 illustrate the connection scheme used to connect the multiplexers of one tile with the LUT's and multiplexers of other tiles.

FIG. 46 illustrates a possible physical architecture of the configurable IC illustrated inFIG. 40.

FIG. 47 illustrates a portion of a configurable IC of some embodiments of the invention.

FIG. 48 illustrates a detailed example of configuration data that configure the circuits to perform particular operations.

FIG. 49 illustrates a system on chip (“SoC”) implementation of a configurable IC.

FIG. 50 illustrates a system in package (“SiP”) implementation for a configurable IC.

FIG. 51 conceptually illustrates a more detailed example of a computing system that has an IC, which includes one of the invention's configurable circuit arrangements that were described above.

DETAILED DESCRIPTION OF THE INVENTION

In the following description, numerous details are set forth for purpose of explanation. However, one of ordinary skill in the art will realize that the invention may be practiced without the use of these specific details. For instance, not all embodiments of the invention need to be practiced with the specific number of bits and/or specific devices (e.g., multiplexers) referred to below. In other instances, well-known structures and devices are shown in block diagram form in order not to obscure the description of the invention with unnecessary detail.

I. Terms and Concepts

A configurable IC is an IC that has configurable circuits. In some embodiments, a configurable IC includes configurable computational units (e.g., configurable logic circuits) and configurable routing circuits for routing the signals to and from the configurable computation units. In addition to configurable circuits, a configurable IC also typically includes non-configurable circuits (e.g., non-configurable logic circuits, interconnect circuits, memories, etc.).

A configurable circuit is a circuit that can “configurably” perform a set of operations. Specifically, a configurable circuit receives “configuration data” that specifies the operation that the configurable circuit has to perform in the set of operations that it can perform. In some embodiments, configuration data is generated outside of the configurable IC. In these embodiments, a set of software tools typically converts a high-level IC design (e.g., a circuit representation or a hardware description language design) into a set of configuration data that can configure the configurable IC (or more accurately, the configurable IC's configurable circuits) to implement the IC design.

Examples of configurable circuits include configurable interconnect circuits and configurable logic circuits. An interconnect circuit is a circuit that can connect an input set to an output set in a variety of manners.FIG. 1 presents a block diagram of a controllable interconnect circuit. Thiscircuit100 has a set ofinput terminals105, a set ofoutput terminals110, and a set ofcontrol terminals115. Theinterconnect circuit100 receives control data along itscontrol terminals115 that causes the interconnect circuit to connect its input terminal set105 to its output terminal set110 in a particular manner.

An interconnect circuit can connect two terminals or pass a signal from one terminal to another by establishing an electrical path between the terminals. Alternatively, an interconnect circuit can establish a connection or pass a signal between two terminals by having the value of a signal that appears at one terminal appear at the other terminal. In connecting two terminals or passing a signal between two terminals, an interconnect circuit in some embodiments might invert the signal (i.e., might have the signal appearing at one terminal inverted by the time it appears at the other terminal). In other words, the interconnect circuit of some embodiments implements a logic inversion, operation in conjunction to its connection operation. Other embodiments, however, do not build such an inversion operation in some or all of their interconnect circuits.

A multiplexer is one example of an interconnect circuit. A multiplexer is a device that has k inputs, n outputs, and s select lines. The select lines typically direct the multiplexer to output n of its k inputs along its n outputs.FIG. 2 illustrates an example of a multiplexer. This example is a four-to-onemultiplexer200 that connects one of its fourinput lines105 to itsoutput line205 based on the control signals115 that it receives along its two select lines.

A configurable interconnect circuit is an interconnect circuit that can configurably perform a set of connection operations. In some embodiments, a configurable interconnect circuit only receives configuration data along its set of control lines. Also, in some embodiments, all of the configurable interconnect circuit's control lines are permanently tied to configuration data. Specifically, in these embodiments, the configurable interconnect circuit's control lines receive configuration data either directly from the storage elements that store the configuration data, or through an indirect connection to these storage elements via one or more interconnect circuits that do not receive configuration data along their control terminals.

FIG. 3 illustrates an example of a configurable four-to-onemultiplexer300 that receives configuration data from a set of storage elements320 (e.g., a set of memory cells, such as SRAM cells). Thismultiplexer300 connects one of its fourinput terminals105 to itsoutput terminal205 based on theconfiguration data115 that is output from the storage elements320. In other words, the configuration data specify how theinterconnect circuit300 should connect the input terminal set105 to theoutput terminal205.

Themultiplexer300 is said to be configurable, as the configuration data set “configures” this multiplexer to use a particular connection scheme that connects its input and output terminal sets in a desired manner. Examples of configurable interconnect circuits can be found in Architecture and CAD for Deep-Submicron FPGAs, Betz, et al., ISBN 0792384601, 1999; and Design of Interconnection Networks for Programmable Logic, Lemieux, et al., ISBN 1-4020-7700-9, 2003. Other examples of configurable interconnect circuits can be found in U.S. patent application Ser. No. 10/882,583, entitled “Configurable Circuits, IC's, and Systems,” filed on Jun. 30, 2004. This application is incorporated in the present application by reference.

A logic circuit is a circuit that can perform a function on a set of input data that it receives. A configurable logic circuit is a logic circuit that can configurably perform a set of functions. In some embodiments, a configurable logic circuit only receives configuration data along its set of control lines. In some embodiments, all of the configurable logic circuit's control lines are permanently tied to configuration data. Specifically, in these embodiments, the configurable logic circuit's control lines receive configuration data either directly from the storage elements that store the configuration data, or through an indirect connection to these storage elements via one or more interconnect circuits that do not receive configuration data along their control lines.

FIG. 4 illustrates a configurable logic circuit. As shown in this figure, aconfigurable logic circuit400 is a logic circuit that receives configuration data along its set ofcontrol lines410. In this example, the configurable logic circuit receives configuration data from a set of storage elements (e.g., SRAM's)415. Thelogic circuit400 is said to be configurable, as the configuration data set “configures” the logic circuit to perform a particular function (from its set of functions) on the input data received along itsinput terminals405 to produce its output data (that is provided along its output terminal set420). Thelogic circuit400 can be reconfigured by writing new data in thestorage elements415.

Some examples of logic circuits include look-up tables (LUT's), universal logic modules (ULM's), sub-ULM's, multiplexers, and PAL's/PLA's. In addition, logic circuits can be complex logic circuits formed by multiple logic and interconnect circuits. Examples of simple and complex logic circuits can be found in Architecture and CAD for Deep-Submicron FPGAs, Betz, et al., ISBN 0792384601, 1999; and Design of Interconnection Networks for Programmable Logic, Lemieux, et al., ISBN 1-4020-7700-9, 2003. Other examples of configurable logic circuits are provided in the above-incorporated U.S. patent application Ser. No. 10/882,583.

As further described below, multiplexers can be used as the logic circuits of some embodiments of the invention. Also, some embodiments use non-configurable logic circuits (i.e., logic circuits that do not receive configuration data or any other control signals). These logic circuits are simple computation circuits that perform a particular computation on an input data set.

A volatile configurable circuit is a configurable circuit that directly or indirectly receives its configuration data from a volatile memory storage. The volatile storage is typically part of the configurable circuit or it is placed adjacent to the configurable circuit within a circuit array that contains the configurable circuit. For instance, themultiplexer300 is a volatile configurable multiplexer as its configuration data is stored in two volatile storage elements320 of the multiplexer. Similarly, thelogic circuit400 is a volatile circuit as its configuration data is stored in its fourstorage elements415. However, the volatile storage does not need to be part of or adjacent to the configurable circuit.

A user-design signal within a configurable IC is a signal that is generated by a circuit (e.g., logic circuit) of the configurable IC. The word “user” in the term “user-design signal” connotes that the signal is a signal that the configurable IC generates for a particular application that a user has configured the IC to perform. User-design signal is abbreviated to user signal in some of the discussion below.

In some embodiments, a user signal is not a configuration or clock signal that is generated by or supplied to the configurable IC. In some embodiments, a user signal is a signal that is a function of at least a portion of the configuration data received by the configurable IC and at least a portion of the inputs to the configurable IC. In these embodiments, the user signal can also be dependent on (i.e., can also be a function of) the state of the configurable IC. The initial state of a configurable IC is a function of the configuration data received by the configurable IC and the inputs to the configurable IC. Subsequent states of the configurable IC are functions of the configuration data received by the configurable IC, the inputs to the configurable IC, and the prior states of the configurable IC.

A circuit array is an array with several circuit elements that are arranged in several rows and columns. One example of a circuit array is a configurable node array, which is an array formed by several configurable circuits (e.g., configurable logic and/or interconnect circuits) arranged in rows and columns. In some cases, all the circuit elements in a configurable node array are configurable circuit elements. For instance,FIG. 5 illustrates an example of aconfigurable node array500 that includes 208configurable nodes505 that are arranged in 13 rows and 16 columns. Each configurable node in such a configurable node array is a configurable circuit that includes one or more configurable sub-circuits. In other cases, the circuit elements in a configurable node array also include non-configurable circuit elements.

In some embodiments, each configurable node in a configurable node array is a simple or complex configurable logic circuit. In some embodiments, each configurable node in a configurable node array is a configurable interconnect circuit. In such an array, a configurable node (i.e., a configurable interconnect circuit) can connect to one or more logic circuits. In turn, such logic circuits in some embodiments might be arranged in terms of another configurable logic-circuit array that is interspersed among the configurable interconnect-circuit array.

Some embodiments might organize the configurable circuits in an arrangement that does not have all the circuits organized in an array with several aligned rows and columns. Accordingly, instead of referring to configurable circuit arrays, the discussion below refers to configurable circuit arrangements. Some arrangements may have configurable circuits arranged in one or more arrays, while other arrangements may not have the configurable circuits arranged in an array.

In some embodiments, some or all configurable circuits in a circuit arrangement have the same or similar circuit structure. For instance, in some embodiments, some or all the circuits have the exact same circuit elements (e.g., have the same set of logic gates and circuit blocks and/or same interconnect circuits), where one or more of these identical elements are configurable elements. One such example would be a set of circuits positioned in an arrangement, where each circuit is formed by a particular set of logic and interconnect circuits. Having circuits with the same circuit elements simplifies the process for designing and fabricating the IC, as it allows the same circuit designs and mask patterns to be repetitively used to design and fabricate the IC.

In some embodiments, the similar configurable circuits not only have the same circuit elements but also have the same exact internal wiring between their circuit elements. For instance, in some embodiments, a particular set of logic and interconnect circuits that are wired in a particular manner forms each circuit in a set of configurable circuits that are in a configurable circuit arrangement. Having such circuits further simplifies the design and fabrication processes as it further simplifies the design and mask making processes.

Also, some embodiments use a circuit arrangement that includes numerous configurable and non-configurable circuits that are placed in multiple rows and columns. In addition, within the above described circuit arrangements (e.g., arrays), some embodiments disperse other circuits (e.g., memory blocks, macro blocks, etc.).

Several figures below illustrate several direct connections between circuits in a circuit arrangement. A direct connection between two circuits in an circuit arrangement is an electrical connection between the two circuits that is achieved by (1) a set of wire segments that traverse through a set of the wiring layers of the IC, and (2) a set of vias when two or more wiring layers are involved.

In some embodiments, a connection between two circuits in a circuit arrangement might also include a set of buffer circuits in some cases. In other words, two circuits are connected in some embodiments by a set of wire segments that possibly traverse through a set of buffer circuits and a set of vias. Buffer circuits are not interconnect circuits or configurable logic circuits. In some embodiments, buffer circuits are part of some or all connections. Buffer circuits might be used to achieve one or more objectives (e.g., maintain the signal strength, reduce noise, alter signal delay, etc.) along the wire segments that establish the direct connections. Inverting buffer circuits may also allow an IC design to reconfigure logic circuits less frequently and/or use fewer types of logic circuits. In some embodiments, buffer circuits are formed by one or more inverters (e.g., two or more inverters that are connected in series).

FIGS. 6 and 7 illustrate examples of two connections, each between two circuits in an arrangement. Each of these connections has one or more intervening buffer circuits. Specifically,FIG. 6 illustrates an example of aconnection615 between two

circuits

605 and610. As shown in this figure, this connection has an interveningbuffer circuit620. In some embodiments, thebuffer circuit620 is an inverter. Accordingly, in these embodiments, theconnection615 inverts a signal supplied by onecircuit605 to theother circuit610.

FIG. 7 illustrates an example of aconnection715 between two

circuits

705 and710. As shown in this figure, thisconnection715 has two intervening

buffer circuits

720 and725. In some embodiments, the

buffer circuits

720 and725 are inverters. Hence, in these embodiments, theconnection715 does not invert a signal supplied bycircuit705 to theother circuit710.

Alternatively, the intermediate buffer circuits between the logic and/or interconnect circuits can be viewed as a part of the devices illustrated in these figures. For instance, the inverters that can be placed after the

devices

705 and710 can be viewed as being part of these devices. Some embodiments use such inverters in order to allow an IC design to reconfigure logic circuits less frequently and/or use fewer types of logic circuits Several figures below “topologically” illustrate several direct connections between circuits in an arrangement. A topological illustration is an illustration that is only meant to show a direct connection between two circuits without specifying a particular geometric layout for the wire segments that establish the direct connection.

II. UMUX

Some embodiments are configurable IC's that have “UMUX's”. A UMUX is a multiplexer that receives user-design signals for at least one of its data inputs and one of its select inputs. A UMUX might receive a user-design signal directly from a configurable logic circuit or indirectly through one or more intermediate configurable interconnect circuits.

Two kinds of UMUX's are full UMUX's and hybrid UMUX's. A full UMUX (or FUMUX) is a UMUX that receives user-design signals for all of its select signals.FIG. 8 illustrates afull UMUX800. This UMUX has a set ofinput terminals805, anoutput terminal810, and a set ofselect terminals815. TheUMUX800 is a four-to-one multiplexer that connects one of its fourdata inputs805 to itsdata output810 based on the value of the two user-signals that it receives along itsselect terminals815.

A hybrid UMUX (or HUMUX) is a UMUX that can receive user-design signals, configuration data, or both user-design signals and configuration data for its select signals.FIG. 9 illustrates anHUMUX900. This HUMUX includes two two-to-one multiplexers920, a four-to-onemultiplexer925, a set ofinput terminals905, anoutput terminal910, and a set ofselect terminals915. From the outside, the HUMUX looks like a four-to-one multiplexer that has fourdata inputs905, onedata output910, and fourselect terminals915. Also, from the outside, the HUMUX looks like it passes one of its fourdata inputs905 to its onedata output910 based on the value of two of the four signals that it receives along its fourselect lines915.

Internally, the two two-to-one multiplexers920 pass two of the signals from the fourselect lines915 to the twoselect terminals940 of the four-to-onemultiplexer925. As shown inFIG. 9, each two-to-one multiplexer920 receives two input signals, which include one user-design signal and one stored configuration signal stored in astorage element945. Each of the two-to-one multiplexers920 outputs one of the two input signals that it receives based on the configuration bit that it receives along its select line950. AlthoughFIG. 9 illustrates two configuration bits stored in two storage elements, other embodiments might drive both multiplexers920 off one configuration bit that is stored in one storage element.

The two signals output by the two multiplexers920 then serve as the select signals of themultiplexer925, and thereby direct this multiplexer to output online910 one of the four input signals that it receives onlines905. The two multiplexers920 can output onlines940 either two user-design signals, two configuration signals, or one user-design signal and one configuration signal. Accordingly, through the two multiplexers920, the operation of themultiplexer925 can be controlled by two user-design signals, two configuration signals, or a mix of user/configuration signals.

References to UMUX's in the description below should be construed to cover either a FUMUX or an HUMUX.

III. UMUX Architectures

Different embodiments of the invention use UMUX's differently in different configurable IC architectures. Some embodiments have a number of UMUX's dispersed between the circuits of a circuit arrangement.FIGS. 10 and 11 illustrate two such embodiments.FIG. 10 illustrates that some embodiments place the UMUX's1005 between thecircuits1010 of acircuit arrangement1000 according to no particular pattern.FIG. 11, on the other hand, illustrates that some embodiments place UMUX's1105 between thecircuits1110 according to a particular pattern that repeats across theentire circuit arrangement1100 or a portion of the arrangement. InFIG. 11, the pattern that is repeated is one that is formed by five UMUX's1105a-1105e.

The circuits (e.g.,1110 or1010) in such circuit arrangements are all logic circuits in some embodiments, while they are interconnect circuits in other embodiments. In yet other embodiments, the circuits (e.g.,1110 or1010) in the circuit arrangement are both logic and interconnect circuits. Also, in some embodiments, the circuit arrangement is a configurable circuit arrangement. Thus, in these embodiments, all the circuits in the circuit arrangement are configurable circuits (i.e., all the circuits are configurable logic circuits, configurable interconnect circuits, or a mixture of configurable logic and interconnect circuits). In some of these embodiments, all or some of the configurable circuits are configurable volatile circuits (i.e., they are configurable volatile logic and/or interconnect circuits).

In the embodiments that disperse UMUX's between the circuits of a circuit arrangement, the UMUX's can be arranged in one or more arrangements.FIG. 12 illustrates an example where a number of UMUX's1205, that are dispersed between thecircuits1210 of a circuit arrangement, form an arrangement of their own.FIG. 13 illustrates an example of arranging UMUX's1305 and1315 in two arrays that are dispersed between thecircuits1310 of a circuit arrangement. Specifically, the UMUX's1305 form one array, while the UMUX's1315 form another array.

Some embodiments position the UMUX's in a circuit arrangement with other types of circuits.FIG. 14 illustrates an example of aUMUX1405 that is positioned in an arrangement of circuit elements1410. This UMUX receives input signals from two

circuit elements

1410aand1410b, and receives select signals from two

circuits elements

1410cand1410d. TheUMUX1405 then outputs a signal tocircuit element1410e. All the circuit elements1410 are configurable circuits (e.g., configurable logic and/or interconnect circuits) in some embodiments. Also, in some embodiments, all these configurable elements are volatile elements. In other embodiments, only some of the circuit elements1410 are volatile configurable circuit elements, while other circuit elements are non-volatile configurable circuit elements or non-configurable circuit elements.

UMUX's can receive user-design signals directly from logic circuits, or indirectly from logic circuits through intervening interconnect circuits.FIG. 15 illustrates a circuit arrangement havinglogic circuits1510,interconnect circuits1505, and UMUX's1515. In this example, aUMUX1515areceives a user-design signal, as a select signal, directly fromlogic circuit1510a, and receives a user-design signal, as an input signal, directly from thelogic circuit1510b. This figure also illustrates aUMUX1515bthat receives a user-design signal as an input signal fromlogic circuit1510dthrough intervening

interconnect circuits

1505band1505c. It also illustrates thisUMUX1515breceiving a user-design signal as a select signal fromlogic circuit1510cthroughinterconnect circuit1505a. Finally, this figure illustrates aUMUX1515cthat receives a first user-design signal directly from alogic circuit1510d, and a second user-design signal indirectly from thelogic circuit1510bthrough theinterconnect circuit1505d. In some embodiments, the intervening

interconnect circuits

1505a,1505b,1505c, and1505dare volatile configurable interconnect circuits.

In the examples illustrated inFIGS. 10-15, UMUX's are part of or are dispersed within a large circuit arrangement that has numerous circuits (e.g., tens, hundreds, thousands, etc. of circuits). In other words, these examples illustrate the use of UMUX's at a macrolevel in configurable IC's. However, in some embodiments, UMUX's are used at the microlevel in a circuit arrangement (e.g., within a circuit or circuits of a, large circuit or circuit arrangement) of a configurable IC. For instance, in some embodiments, the circuits in a configurable IC's circuit arrangement include complex logic circuits that are formed by multiple logic and interconnect circuits. In such embodiments, complex logic circuits can be formed by one or more UMUX's.

FIG. 16 illustrates an example of such an embodiment. Specifically, this figure illustrates acircuit arrangement1600 that includes a number oflogic circuits1605 arranged in a number of rows and columns. This figure further illustrates an explodedview1610 of one of thelogic circuits1605a. As shown in this view, thelogic circuit1605ais formed by four UMUX's1615 and oneinterconnect circuit1620. Each of the UMUX's receives input signals and user-design signals, which direct the UMUX to output one of its input signals. The UMUX output signals are provided to theconfigurable interconnect circuit1620, which then outputs one or more of these output signals as the output of thelogic circuit1605a.

In some embodiments, all thelogic circuits1605 in thearray1600 have the structure of thelogic circuit1605a. In other embodiments, only some of thelogic circuits1605 have this structure, while other logic circuits have a different structure. For instance, some embodiments use configurable logic circuits in combination with the UMUX-based logic circuits (i.e., with logic circuits that are similar to thecircuit1605a). Other embodiments use non-configurable logic circuits with the UMUX-based logic circuits. Yet other embodiments use a combination of configurable logic circuits, non-configurable logic circuits, and UMUX-based logic circuits.

IV. Applications

A. Supering

Some embodiments use UMUX's for “supering” multiple logic circuits. Supering refers to combining the outputs of multiple logic circuits to form a more complex logic circuit.FIG. 17 illustrates one such embodiment. This figure illustrates a macro-level view of aconfigurable circuit arrangement1700 with numerous (e.g., hundreds, thousands, etc.) circuits arranged in rows and columns. This array includes numerous configurablevolatile logic circuits1705, configurablevolatile interconnect circuits1710, and UMUX's1715.

As shown inFIG. 17, oneUMUX1715areceives as input signals the outputs of two

logic circuits

1705aand1705bthrough two intervening

configurable interconnect circuits

1710aand1710b. ThisUMUX1715aalso receives as a select signal the output of alogic circuit1705cthrough an interveningconfigurable interconnect circuit1710c, which is a volatile configurable circuit in some embodiments. Based on the received select signal, theUMUX1715aperforms a function on its received input signals. The UMUX performs this function by outputting one of its input signals along its output line based on the value of the received select signal.

The circuit structure illustrated inFIG. 17 can be used to implement a four-input LUT with two three-input LUT's and a UMUX. In other words, if the

logic circuits

1705aand1705bare three-input LUT's, theUMUX1715acan be used to create a “super” four-input LUT by combining the outputs of three-input LUT's1705aand1705bbased on the output of thelogic circuit1705c. In this case, the supered four-input LUT is the circuit formed by LUT's1705aand1705b, theinterconnect circuit1710a, and theUMUX1715a.

B. Memory

Some embodiments of the invention use UMUX's to implement configurable circuit arrangements with configurable memory read and write operations. Some configurable IC's today have configurable circuit arrangements with configurable memory structures.FIG. 18 illustrates one such priorconfigurable circuit arrangement1800. This circuit arrangement includes numerousconfigurable circuits1805 that are arranged in rows and columns. It also includes twoconfigurable memories1810. Each of these two memories includes amemory array1820, i/o circuitry1825, and configuration-storingstorage elements1830.

The configuration-storingstorage elements1830 store configuration data. This data configures each memory to input and/or output a certain number of bits. Specifically, thememory array1820 receives an address along itsaddress line1835. During a write operation, the address identifies a location in the memory to which the i/o circuitry needs to write a fixed number of bits. Alternatively, during a read operation, this address identifies a location in the memory from which the i/o circuitry needs to read a fixed number of bits (e.g., 8 bits).

The i/o circuitry1825 receives configuration data from thestorage elements1830. Based on the received configuration data, the i/o circuitry then outputs a sub-set of the data bits (e.g., outputs 4 bits) that it receives from thememory array1820 during a read operation. In other words, the number of bits that the i/o circuitry might output from a location in the memory array depends on the configuration data stored in the configuration-storingstorage elements1830.

Requiring thememories1810 to haveconfiguration storage elements1830 and/or circuitry responsive to these storage elements complicates the structure of these memories, and thereby increases their size. Accordingly, to use simpler memory structures in a configurable circuit arrangement of a configurable IC, some embodiments use UMUX's to implement configurable circuit arrangements with configurable memory read operations.

FIG. 19 illustrates one such embodiment. This figure illustrates aconfigurable circuit arrangement1900 that includes numerousconfigurable interconnect circuits1905, UMUX's1910, andlogic circuits1920. Theinterconnect circuits1905 are arranged in an arrangement with several rows and columns. Dispersed within this arrangement of interconnect circuits are the logic circuits, which are also arranged in an arrangement with multiple rows and columns. The UMUX's are dispersed within these arrangements.Several memories1915 are also dispersed within these arrangements. In some embodiments, these memories all have the same width. For instance, in some cases, these memories are 256 byte memories that output eight-bit words.

FIG. 19 illustrates two examples for reading out data from anon-configurable memory1915a. The output of each of these memories is always eight-bits wide. To read out a byte of data (i.e., to read out eight bits of data), each memory (e.g.,1915a) receives an eight-bit address that identifies one of 256 locations in the memory.

As shown inFIG. 19, the eight-bit wide output of thememory1915acan be supplied directly to alogic circuit1920athrough aninterconnect circuit1905a. Alternatively, as shown in this figure, a four-bit wide output of thememory1915acan be supplied to thelogic circuit1920bthrough aUMUX1910a. Specifically, through theinterconnect circuit1905a, the eight-bit wide output can be supplied to theUMUX1910a. TheUMUX1910ais an 8×4 multiplexer that outputs either its bottom four input signals or its top four input signals to alogic circuit1920b. Specifically, theUMUX1910areceives asignal1925 on its select line. Based on the value of this signal, the UMUX then selects its bottom four input signals or its top four input signals to output.

Thebit1925 that theUMUX1910areceives can be viewed as a ninth address bit that helps identify and output a 4-bit word from thememory1915a. In other words, theUMUX1910aallows the 256×8 memory, which is accessed with an eight-bit address, to be treated as a 512×4 memory that is accessible with a nine-bit address.

FIG. 20 illustrates an example of using UMUX's to write data to anon-configurable memory array2015 in aconfigurable circuit arrangement2000. In this figure, the memory array has a set2005 of four UMUX's to the right of the memory array. In the example illustrated inFIG. 20, the UMUX set2005 is used for writing data to the memory array. The output of the UMUX set2005 connects to the memory array through a set of configurable interconnects2025, while the input of the UMUX set2005 connects to the memory array through a set of configurable interconnects2020.

During a write operation, an address is supplied to thememory array2015 that identifies a four-bit location in this array. This four-bit location is a location to which the memory array will write four bits that are supplied through the UMUX set2005 via the interconnect set2025. Each UMUX in the UMUX set2005 receives two bits. During a write operation, one bit comes from the user design, while the other bit comes from the memory. Specifically, during a write operation, thememory array2015 provides the four bits currently stored in the location identified by the received address to the UMUX set2005 through the interconnect set2020.

During the write operation, each of the UMUX's2005 receives a user signal on its select line that specifies, which of its two inputs (the user design input or the feedback, memory input) it should output. Through configurable interconnect set2025, the outputs of the UMUX's are then provided to thememory array2015 as data to write to the identified four-bit location.

The advantage of the UMUX-driven approach for accessing memories is that it allows all the memories in a configurable circuit arrangement to have the adjustable widths without requiring the use of configuration data. For instance, in the example illustrated inFIG. 20, the UMUX set2005 allows two bits to be written in a memory that always reads and writes on a four-bit basis. Specifically, by providing one additional address bit that is decoded to provide two UMUX select signals, two UMUX provide two new bits to the memory while the other two UMUX write back to the memory two bits that it currently stores.

The UMUX-driven approach also allows the memories to have a relatively simple structure as these memories no longer need to store configuration data or circuitry responsive to configuration data. Moreover, the UMUX-driven approach allows the user to effectively have different memory operations with simpler memory structures simply through the programmable operation of the IC.

C. Shifter

Some embodiments of the invention use UMUX's to implement circuit arrangements with configurable shift operations.FIG. 21 illustrates one such embodiment. Specifically, this figure illustrates acircuit arrangement2100 that has numerous circuits arranged in numerous rows and columns. These circuits include volatile

configurable interconnect circuits

2105,2110, and2115, and

logic circuits

2120,2125 and2130.

Thelogic circuit2120 includes four four-to-one UMUX's2135. Each of these UMUX's serves as a shifter. These UMUX's can be used to seamlessly perform a variable shift operation whenever the output of a logic circuit in thearrangement2100 needs to be shifted by a certain number of bits that is to be determined by the output of another logic circuit. All that needs to be done is to supply to the UMUX's through configurable interconnect circuits (1) the bits that need to be shifted, and (2) the bits that need to specify the amount of the shift.

FIG. 21 illustrates one example of configurably connecting the UMUX's2135 to logic circuits in thecircuit arrangement2100. Specifically, this figure illustrates the supplying of the output oflogic circuit2125 to the UMUX's2135 through theconfigurable interconnect circuit2105. In this example, the output of thelogic circuit2125 is four bits wide. Also, in this example, eachUMUX2135 receives the four-bit output of thelogic circuit2125 in a different order, as shown inFIG. 21.

FIG. 21 also illustrates the supplying of the output of thelogic circuit2130 to the UMUX's2135 through theconfigurable interconnect circuit2110. In this example, the output of thelogic circuit2130 is two-bits wide. The value of these two bits cause each of the UMUX's2135 to output a particular one of their four input bits. This output operation of the UMUX's2135 can result in a shift of the output of thelogic circuit2125. For instance, lets assume that thelogic circuit2130 output is a binary value of 10. For the order of the inputs of the UMUX's2135 illustrated inFIG. 21, this output fromlogic circuit2130 causes the UMUX's2135 to output2301 (where 0, 1, 2, 3 refer to bits in the four bit output from thelogic circuit2125 and the order of that these bits are received is 0123).

D. Interconnect Circuits

An HUMUX blurs the line between a communication circuit and a computation circuit. In a configurable IC that can be reconfigured in real time (i.e., in a reconfigurable IC), an HUMUX can serve as a logic circuit during some clock cycles, while serving as an interconnect circuit during other clock cycles. This ability is advantageous in a variety of contexts.

1. Hybrid Routing Mux and Logic Circuit

An HUMUX can be used as a hybrid routing multiplexer and logic circuit in a circuit arrangement of a configurable IC. In a circuit arrangement, a routing multiplexer is an interconnect circuit that connects other logic and/or interconnect circuits in the arrangement. In some embodiments, a routing multiplexer either provides its output to several logic and/or interconnect circuits (i.e., has a fan out greater than 1), or provides its output to other interconnect circuits. An HUMUX acts as a routing mux (i.e., does not perform any computation and only serves to relay signals from other circuits) when its operation is completely controlled by configuration data. On the other hand, an HUMUX serves as a logic circuit when its operation is not completely controlled by configuration data.

FIG. 22 illustrates an example of acircuit arrangement2200 that includesnumerous logic circuits2205,interconnect circuits2210, and HUMUX's2215. In some embodiments, some or all of the logic circuits and/or interconnect circuits are configurable circuits. The above-describedHUMUX900 ofFIG. 9 is an example of an HUMUX that can be used as the HUMUX's2215. TheHUMUX900 acts as a four-to-one multiplexer that passes one of its fourdata inputs905 to its onedata output910 based on the value of two of the four signals that it receives along its fourselect lines915. TheHUMUX900 receives two user-design signals and two configuration data bits on its four select lines. The four-to-two multiplexer920 of this HUMUX passes two of these four signals to the twoselect terminals940 of the four-to-onemultiplexer925 of the HUMUX, which, in turn, identify one of the fourinput signals905 for themultiplexer925 to output.

Accordingly, the HUMUX's operation (1) is completely controlled by the configuration data bits when the two signals supplied to theselect terminals940 are the configuration data bits, (2) is completely controlled by the user-design signals when the two signals supplied to theselect terminals940 are users signals, and (3) is partially controlled by the user-design signals when the two signals supplied to theselect terminals940 include one user-design signal and one configuration data bit.

As mentioned above, an HUMUX acts as a routing multiplexer when its operation is completely controlled by configuration data; otherwise, the HUMUX serves as a logic circuit. Hence, when theHUMUX900 is used as anHUMUX2215, theHUMUX2215/900 acts as a routing multiplexer when the two signals supplied to theselect terminals940 are the configuration data bits stored in theconfiguration data elements945. Otherwise, when at least one of the two signals supplied to theselect terminals940 is a user-design signal, theHUMUX900 acts as a logic circuit.

It is advantageous to use HUMUX's in a circuit arrangement of a configurable IC, because they can be used to augment the routing resources in some cycles, while augmenting the computational resources in other cycles. Having such dual-use devices not only increases the flexibility of the design of the configurable IC, but also reduces the size of the IC by requiring fewer overall number of logic and interconnect circuits.

2. Hybrid Input-Select Multiplexer and Logic Circuit

An HUMUX can also be used as a hybrid input-select multiplexer and logic circuit in a circuit arrangement of a configurable IC. In a circuit arrangement, an input-select multiplexer is an interconnect circuit associated with a particular logic circuit in the arrangement. This interconnect circuit receives several input signals and passes a sub-set (e.g., one) of these input signals to its corresponding logic circuit. Accordingly, in some embodiments, an input-select multiplexer provides its output to only one logic circuit (i.e., has a fan out of 1), although this is not the case in all embodiments, as further discussed below.

One example of input-select multiplexers is presented inFIG. 23. This figure illustrates an example of acircuit arrangement2300 withnumerous logic circuits2305 androuting interconnect circuits2310. Theinterconnect circuits2310 are responsible for relaying signals between logic circuits and/or other interconnect circuits. The logic circuits, on the other hand, are responsible for performing computations on their respective input signals. In some embodiments, some or all the interconnect and/or logic are configurable circuits.

As shown inFIG. 23, alogic circuit2305aincludes three input-select multiplexers2315 and acore logic circuit2320. The three input-select multiplexers2315 receive input signals for thecore logic circuit2320. When operating as input-select multiplexers, these multiplexers pass a sub-set of the received input signals to thecore logic circuit2320, which performs a function on these input signals.

HUMUX's can be used for one or more of the input-select multiplexers2315. Some embodiments have an HUMUX for only one of the input-select multiplexers2315. An HUMUX acts as an input-select multiplexer (i.e., does not perform any computation and only serves to relay a signal to its associated logic circuit) when its operation is completely controlled by configuration data. For instance, when theHUMUX900 ofFIG. 9 is used as theHUMUX2315a, theHUMUX900 acts as an input-select multiplexer when the two signals supplied to theselect terminals940 are the configuration data bits stored in theconfiguration storage elements945.

Otherwise, when the operation of the HUMUX is not completely controlled by configuration data, theHUMUX2315aacts as a logic circuit. For instance, when theHUMUX900 is used as anHUMUX2315a, theHUMUX900 acts as a logic circuit when at least one of the two signals supplied to theselect terminals940 is a user-design signal.

When such an HUMUX (i.e., an HUMUX at the input of a logic circuit) acts as a logic circuit, the HUMUX augments the computational abilities of the logic circuit. For instance, in the example illustrated inFIG. 23, lets assume that thecore logic circuit2320 is a three-input LUT that receives three inputs from three multiplexers2315. Lets further assume that the four-to-one HUMUX900 is used as theHUMUX2315aat the input of thecore logic circuit2320. This HUMUX receives two user-

signal bits

2330 and2335 that would select one of the fourinputs2340 for output to the three-input LUT2320.

The combination of the logic circuit formed by the three-input LUT2320 and theHUMUX2315a(acting as a logic circuit) serves as a four-input logic circuit. It is expected that such a four-input logic circuit would be capable of implementing as many as 50 of the 220 most prevalent four-input functions that a configurable logic circuit can perform. Such added functionality comes relatively inexpensively, as it only comes at the expense of the slightly larger size of the HUMUX as the input select multiplexer2315. Also, such added functionality not only increases the flexibility of the design of the configurable IC, but also might reduce the overall size of the IC by requiring fewer overall number of logic circuits.

In some embodiments, an input-select multiplexer can have a fan out of greater than 1. In these embodiments, the input-select multiplexer still connects directly to one logic circuit so that it can select an input of this logic circuit. However, in these embodiments, the input-select multiplexer also can provide its output to other circuits in the circuit arrangement.

3. HUMUX's as Input-Select Multiplexers, Routing Muxes, and Logic Circuits

The IC's of some embodiments have circuit arrangements that use HUMUX's as both input-select multiplexers and routing multiplexers.FIG. 24 illustrates one such embodiment. Specifically, this figure illustrates acircuit arrangement2400 that hasnumerous logic circuits2405 andinterconnect circuits2407 arranged in numerous rows and columns. Thiscircuit arrangement2400 also includes several HUMUX's asrouting interconnect circuits2410 and several HUMUX's2415aas inputselect multiplexers2415 of somelogic circuits2405a.

It is advantageous to use in a configurable IC HUMUX's as both input select multiplexers and routing interconnect multiplexers. This is because such HUMUX's allow the configurable IC to form very complex logic circuit structures relatively easily by supering logic and HUMUX circuits. Some embodiments, however, use HUMUX's only for input select multiplexers.

V. HUMUX Useful for Decomposing all Functions

HUMUX's are hybrid interconnect/logic circuits. In other words, as mentioned above, HUMUX's can serve as logic and interconnect circuits in a configurable IC. This hybrid quality is especially advantageous since, as logic circuits, HUMUX's can be used to decompose and implement functions. In order to decompose and implement functions with HUMUX's, some embodiments define one input of some or all HUMUX's to be a permanently inverting input.

More specifically, some embodiments use Shannon and Davio decomposition to decompose functions. Some embodiments use the multiplexer function of the HUMUX to implement the “if-then-else” (ITE) expression needed for a Shannon decomposition. Through this ITE expression, these embodiments implement the AND function and its NPN equivalents, where NPN stands for negate input, permute input, and negate output.

FIGS. 25-30 illustrate how the multiplexer functionality of an HUMUX allows the HUMUX to implement the AND function and its NPN equivalents. Specifically,FIG. 25 illustrates a two-input multiplexer2500 that receives variables “b” and “c” on its input lines and variable “a” on its select line. As shown in this figure, thismultiplexer2500 has an output (i.e., performs a function) that can be expressed as [(a·b)+(ā·c)].

Themultiplexer2500 can be used to generate the AND function and its NPN equivalents by routing constants to one of the multiplexer inputs and/or performing NPN operations. For instance,FIG. 26 illustrates that the multiplexer can perform the AND function by routing a binary 0 to its second input. Similarly,FIG. 27 illustrates that routing a binary 1 to the multiplexer's first input, allows the multiplexer to perform the OR function, which is one of the NPN equivalent functions of the AND function.

Like Shannon decomposition, Davio decomposition relies on the multiplexer functionality. However, unlike Shannon decomposition, Davio decomposition uses the XOR function to decompose a function.FIG. 28 illustrates a two-input multiplexer2800 that implements an XOR function. As shown in this figure, themultiplexer2800 has two inputs, one of which is permanently inverting. Both inputs of themultiplexer2800 receives the same input signal “b”, while this multiplexer's select line receives the input signal “a”. Because of these input signals and the permanently inverting input, the multiplexer computes [(a·b)+(ā·b)], which is another representation of (a⊕b), i.e., of “a” XOR “b”. Themultiplexer2800 can also compute NPN equivalents of the XOR function through NPN operations (i.e., through negating its inputs, permuting its inputs, and/or negating its output).

The multiplexers illustrated inFIGS. 25-28 can all be the core multiplexer of an HUMUX (e.g., a multiplexer like the four-to-onemultiplexer925 of theHUMUX900 ofFIG. 9). When the HUMUX acts as a logic circuit, this core multiplexer receives a user signal. For instance,FIG. 29 illustrates an example of anHUMUX2900 that has acore multiplexer2905 that operates like themultiplexer2500 ofFIGS. 25-27 as it receives the user signal “a” through theselect multiplexer2910.

FIG. 30 illustrates another example of anHUMUX3000 that has acore multiplexer3005 that operates like themultiplexer2800 ofFIG. 28 as it receives the user signal “a” through theselect multiplexer3010. Like themultiplexer2800, thecore multiplexer3005 has one permanently inverting input. One of ordinary skill will realize that other embodiments might obviate the need for a permanently inverting input in the core multiplexer of an HUMUX by routing inverted inputs to such a multiplexer.

FIG. 31 illustrates arecursive process3100 that a synthesizer can perform to decompose a function. This process uses a series of Shannon and/or Davio decompositions to decompose a function with more than two inputs into one or more functions with two inputs. The synthesizer can then implement each two-input function by the multiplexers illustrated inFIGS. 25-28 and their NPN equivalents.

As shown inFIG. 31, theprocess3100 initially receives (at3105) a function with more than two inputs. It then determines (at3110) whether it should use Shannon or Davio decomposition to remove one variable from the received function, in order to decompose the function. In some embodiments, the process makes this determination by using commonly known techniques, such as sifting on ordered Kronecker Decision Diagrams. For instance, some of these techniques determine whether the received function is highly dependent on a variable that the process can determine (at3110) to remove. If so, the process determines (at3110) to use the Shannon decomposition to remove the variable. Otherwise, the process determines (at3110) to use the Davio decomposition to remove the variable.

When the process determines (at3110) to use the Shannon decomposition to remove the variable, the process transitions to3115. At3115, the process performs Shannon decomposition on the received function. The decomposition operation results in a new expression of the received function that is dependent on the removed variable and two functions that are not dependent on the removed variable. Equation (A) illustrates an example of a Shannon decomposition for a five variable function F(a,b,c,d,e).

F(a,b,c,d,e)=[{a·F₁(b,c,d,e)}+{ā·F₀(b,c,d,e)}] (A)

As expressed in Equation (A), the Shannon decomposition can decompose the five-variable function F(a,b,c,d,e) into two four-variable functions F₁and F₀, which are individually AND'ed with the removed variable “a” and its complement, where the result of the AND operations are OR'ed together. After performing the Shannon decomposition at3115, theprocess3100 transitions to3120, which will be described further below.

When the process determines (at3110) to use the Davio decomposition to remove the variable, the process transitions to3125. At3125, the process performs Davio decomposition on the received function. This decomposition operation results in a new expression of the received function that is dependent on the removed variable and two functions that are not dependent on the removed variable.

There are two variations of Davio decomposition, a positive Davio decomposition and a negative Davio decomposition. Equation (B) illustrates an example of a positive Davio decomposition for the five variable function F(a,b,c,d,e).

F(a,b,c,d,e)=[{a·F₁(b,c,d,e)}⊕{F₂(b,c,d,e)}]. (B)

where

F₂(b,c,d,e)=F₁(b,c,d,e)⊕F₀(b,c,d,e), (C)

where F₀and F₁are identical in Equations (A), (B), and (C). As expressed in Equation (B), the positive Davio decomposition can decompose the five-variable function F(a,b,c,d,e) into two four-variable functions F₁and F₂, with the first function F₁AND'ed with the removed variable “a” and the result XOR'ed with the second function F₂.

Equation (D) illustrates an example of a negative Davio decomposition for the five variable function F(a,b,c,d,e).

F(a,b,c,d,e)=[{a·F₀(b,c,d,e)}⊕{F₂(b,c,d,e)}]. (D)

where F₀and F₂are identical in Equations (A), (B), (C), and (D). As expressed in Equation (D), the negative Davio decomposition can decompose the five-variable function F(a,b,c,d,e) into two four-variable functions F₀and F₂, with the first function F₀AND'ed with the removed variable “a” and the result XOR'ed with the second function F₂.

After performing the Davio decomposition at3125, theprocess3100 transitions to3120. At3120, the process determines whether any of the functions that resulted from the decomposition (e.g., whether functions F₀, F₁, or F₂) has more than two variables. When the process determines (at3120) that none of the resulting functions have more than two variables, the process returns (at3135) the two-input functions that the process identified itself or identified through its recursive calls, and then ends. On the other hand, if the process determines (at3120) that one or more of the resulting functions has more than two variables, the process recursively calls itself (at3130) once for each of these functions. Once the process has received the result of its recursive call, it returns (at3135) the two-input functions that the process identified itself or identified through its recursive calls, and then ends.

When the original function has been broken down into a series of two-input functions, the original function can be implemented by a series of HUMUX's that perform two-input function AND and/or XOR operations and their NPN equivalents. When an HUMUX has more than two input lines and more than one select line, the HUMUX can be reduced into a two input HUMUX by setting all the select lines except one to default values, which, in turn, renders all the inputs of the HUMUX irrelevant except two of its inputs.

Theprocess3100 decomposes a function into a series of two-input functions. It is not necessary, however, for the decomposition operation to decompose the function into a series of two-input functions. The end result of other decomposition processes might include functions with three or more variables. These more complex functions, in turn, can be implemented by using other logic circuits (e.g., three or four input LUT's) or using HUMUX's with more than two inputs. Accordingly, the above-describedprocess3100 presents only one manner of decomposing a function. Thisprocess3100 serves as an example of how a function with more than two variables can be decomposed through Shannon and/or Davio decomposition into a series of two input functions that can be implemented by HUMUX's.

Also, the discussion above focuses on using HUMUX's to decompose and implement complex functions. FUMUX's, however, can also be used to decompose and implement complex functions, by having some FUMUX's with at least one input that is a permanently inverting input or that receives an inverted version of another of its inputs. The advantage of using HUMUX's in an IC, however, is that HUMUX's can serve not only as logic circuits (that can decompose and implement complex functions) but can also server as interconnect circuits.

VI. Sub-Cycle Reconfigurable HUMUX

Reconfigurable IC's are one type of configurable IC's. Reconfigurable IC's are configurable IC's that can reconfigure during runtime. In other words, a reconfigurable IC is an IC that has reconfigurable logic circuits and/or reconfigurable interconnect circuits, where the reconfigurable logic and/or interconnect circuits are configurable logic and/or interconnect circuits that can “reconfigure” more than once at runtime. A configurable logic or interconnect circuit reconfigures when it receives a different set of configuration data. Some embodiments of the invention are implemented in reconfigurable IC's that are sub-cycle reconfigurable (i.e., can reconfigure circuits on a sub-cycle basis).

FIGS. 32-36 illustrate a more detailed example of anHUMUX3200 of some embodiments of the invention. This example is a sub-cycle reconfigurable HUMUX that is implemented in complementary pass logic (CPL) for a sub-cycle reconfigurable IC. In a CPL implementation of a circuit, a complementary pair of signals represents each logic signal, where an empty circle at the input or output of a circuit denotes the complementary input or output of the circuit in the figures. In other words, the circuit receives true and complement sets of input signals and provides true and complement sets of output signals. A sub-cycle reconfigurable IC is an IC that has reconfigurable circuits that can receive configuration data sets on a sub-cycle basis and therefore can reconfigure on a sub-cycle basis. Examples of sub-cycle reconfigurable circuits are disclosed in U.S. patent application entitled “Configurable IC with Interconnect Circuits that also Perform Storage Operations”, which is filed concurrently with the present application, with attorney docket number TBUL.P0022. This application is incorporated herein by reference.

As shown inFIG. 32, theHUMUX3200 is an eight-to-one HUMUX that is formed by four

multiplexers

3205,3210,3215, and3220. The inputs and outputs of these multiplexers are shown as thick lines to indicate that each of these lines represents a CPL true/complement pair of lines.

As shown inFIG. 32, themultiplexer3205 is an eight-to-one multiplexer that, on a sub-cycle basis, connects one of its input lines to its output line based on the values of the signals S2, S1, and S0′, which it receives along its three select lines. In response to three signals ST0, ST1, and CLK (which is not illustrated inFIG. 32), themultiplexer3210 supplies two of the select signals S2 and S1 to themultiplexer3205 on a sub-cycle basis. Specifically, based on the signals ST0 and ST1 that it receives on its select lines, themultiplexer3210 connects one of its four three-bit input lines (each of which connects to astorage element3225 that stores configuration data) to its three output lines. Hence, the three output lines ofmultiplexer3210 provide three configuration select signals S2, S1, and S0. Two of these output lines connect to the third and second select lines of themultiplexer3205, in order to provide the select signals S2 and S1.

The first output line of themultiplexer3210 carries the first select signal S0. This output line connects to one of the two input lines of themultiplexer3220. The other input line of themultiplexer3220 receives a user signal. Through its two input lines, themultiplexer3220 receives two inputs on a sub-cycle basis. Based on the signal that it receives on a sub-cycle basis on its select line from themultiplexer3215, themultiplexer3220 supplies one of its two inputs to its output line. This output line connects to the first select line of themultiplexer3205 to provide the select signal S0′. Hence, the signal S0′ is a signal that in each sub-cycle might be either a user signal or configuration-driven select signal S0.

Which of these signals gets routed to themultiplexer3205 as the select signal S0′ depends on the value of the configuration data output from themultiplexer3215 on a sub-cycle basis. Themultiplexer3215 is a four-to-one multiplexer that (1) has its four inputs connected to four storage elements storing four configuration data bits, and (2) has one output that receives one of the four configuration data bits in each sub-cycle based on the signals ST0 and ST1 supplied to the select lines of themultiplexer3215.

FIG. 33 illustrates amultiplexer3300 that in some embodiments is used as themultiplexer3205. As shown inFIG. 33, themultiplexer3300 includes one set ofinput buffers3305, three sets of

pass transistors

3310,3315, and3320, two pull-up

PMOS transistors

3325 and3330, and two

output buffers

3335 and3340. One subset of the input buffers3305 receives eight input bits (0-7), while another subset of the input buffers3305 receives the complement of the eight input bits (i.e., receives0-7). These input buffers serve to buffer thefirst set3310 of pass transistors.

Thefirst set3310 of pass transistors receive the third select bit S2 or the complement of this bit, while thesecond set3315 of pass transistors receive the second select bit S1 or the complement of this bit. Thethird set3320 of pass transistors receive the first select bit S0′ or its complement. The three select bits S2, S1, and S0 cause the pass transistors to pass one of the input bits and the complement of this input bit to two

intermediate output nodes

3355 and3360 of thecircuit3300. For instance, when the enable signal is low, and the select bits are 011, the

pass transistors

3365a,3370a,3375a, and3365b,3370b, and3375bturn on to pass the 6 and 6 input signals to the

intermediate output nodes

3355 and3360.

The pull-up

PMOS transistors

3325 and3330 are used to pull-up quickly the

intermediate output nodes

3355 and3360, and to regenerate the voltage levels at the nodes that have been degenerated by the NMOS threshold drops, when these nodes need to be at a high voltage. In other words, these pull-up transistors are used because the NMOS pass transistors are slower than PMOS transistors in pulling a node to a high voltage. Thus, for instance, when the 6^thinput signal is high, the enable signal is low, and the select bits are 011, the pass transistors3365-3375 start to pullnode3355 high and to pushnode3360 low. The low voltage onnode3360 turns on the pull-uptransistor3325, which, in turn, accelerates the pull-up ofnode3355.

The

output buffer inverters

3335 and3340 are used to allow thecircuit3300 to drive a load. These buffers are formed by more than one inverter in some embodiments. The outputs of these buffers are the

final output

3380 and3385 of themultiplexer3300. It should be noted that, in some embodiments, the

output buffers

3335 and3340 are followed by multiple inverters.

AlthoughFIG. 32 conceptually illustrates themultiplexer3220 as a separate multiplexer that is after themultiplexer3210, some embodiments implement themultiplexer3220 as part of a multiplexer structure that is used to define themultiplexer3210. In some embodiments, themultiplexer3210 ofFIG. 32 is formed by three multiplexer structures. One structure is used to generate the third select signal S2, the other structure is used to generate the second select signal S1, and the last structure is used to generate the first select signal S0′.

FIG. 34 illustrates a CPL implementation of a two-tier multiplexer structure3400 for generating the second signal S1 and its complement. The multiplexer structure for generating the third select signal S2 is identical to thestructure3400. In addition, an identical circuit can be used to implement themultiplexer3215 ofFIG. 32. Themultiplexer structure3400 is a two-tiered multiplexer structure, where one tier of multiplexers is driven by the signals ST0 and ST1, while the other tier of multiplexers are driven by the clock signal CLK that operates at twice the rate of the signals ST0 and ST1, as further described below.

As illustrated inFIG. 34, the selectsignal generation circuit3400 can be divided into four sections, which are (1)storage cell section3405, (2) first two-to-onemultiplexer section3410, (3) second two-to-onemultiplexer section3415, and (4) pull-upPMOS transistor sections3420. Thefirst section3405 includes four storage cells3425a-3425dthat store four configuration bits for four sub-cycles. In other words, each storage cell provides a configuration bit3430 and the complement of this bit3435, where each such pair of bits provides the select bit signal S1 and its complement during a particular sub-cycle.

The second section includes two

multiplexers

3440 and3445 that are driven by two signals ST0 and ST1 that are offset by 90°, and the differential complementST0 andST1 of these signals. The third section is one two-to-one multiplexer3415 that is driven by the clock signal CLK and its differential complementCLK.

FIG. 35 illustrates an example of the signals CLK, ST0, and ST1. Some embodiments use the multiplexer/storage circuit3300 and the select-signal generator3400 in a configurable IC that implements a design that has a primary clock rate of ×MHZ (e.g., 200 MHZ) through a four sub-cycle implementation that effectively operates at 4×MHZ. In some of these embodiments, the two signals ST0 and ST1 would operate at ×MHZ, while the clock signal CLK would operate at 2×MHZ.

Thefourth section3420 includes two pull-up

PMOS transistors

3485 and3490, which are used to quickly pull-up the output of themultiplexer3415 that is high. The two complementary outputs of themultiplexer3415 provide the select signal S1 and its complement. The S1 select signal and its complement drive the pass transistor set3315 inFIG. 33.

FIG. 34 illustrates onepossible implementation3450 of themultiplexer3445 and the connections of thismultiplexer3445 and the

storage cells

3425cand3425d. As shown in this figure, themultiplexer3445 can be implemented by four pass transistors, where two

transistors

3455 and3460 receive the

true configuration bits

3430cand3430dfrom the third and

fourth storage cells

3425cand3425d, while the other two

transistors

3465 and3470 receive the

complement configuration bits

3435cand3435dfrom the third and fourth storage cells. As further shown,

transistors

3455 and3465 are driven by signal ST1, while

transistors

3460 and3470 are driven by the complementST1 of signal ST1. A similar implementation can be used formultiplexer3440. However, the pass transistors3455-3470 of themultiplexer3440 would be driven by the signal ST0 and its complementST0.

FIG. 34 also illustrates one possible implementation of the two-to-one multiplexer3415. This implementation is similar to theimplementation3450 of themultiplexer3445. However, instead of the signal ST1, the pass transistors3455-3470 of themultiplexer3415 are driven by the CLK andCLK signals. Also, these transistors receive a different set of signals. Specifically, the

transistors

3455 and3465 of themultiplexer3415 receive the true and complement outputs of themultiplexer3440, while the

transistors

3460 and3470 of themultiplexer3415 receive the true and complement outputs of themultiplexer3445.

The

transistors

3455 and3465 of the multiplexer3445 (1) output the true and complement configuration bits stored in thestorage cells3425cwhen the signal ST1 is high, and (2) output the true and complement configuration bits stored in thestorage cells3425dwhen the signal ST1 is low. Similarly, the

transistors

3455 and3465 of the multiplexer3440 (1) output the true and complement configuration bits stored in thestorage cells3425awhen the signal ST0 is high, and (2) output the true and complement configuration bits stored in thestorage cells3425bwhen the signal ST0 is low. Finally, the

transistors

3455 and3465 of the multiplexer3415 (1) output the true and complement output bits of themultiplexer3440 when the clock CLK is high, and (2) output the true and complement output bits of themultiplexer3445 when the clock signal CLK is low.

Given the above-described operations of

multiplexers

3440,3445, and3415, and given the 90° offset between signals ST0 and ST1 and the faster frequency of the clock signal CLK,FIG. 35 illustrates the value of the select signal S1 and its complement that thecircuit3400 generates during each half-cycle of the clock signal CLK. This clocking scheme hides all the timing of the selection of the configuration bits from the storage cells3425 behind the two-to-one multiplexer3415. For instance, while themultiplexer3440 is switching between outputting the configuration bits stored incell3425aand the bits stored incell3425b, the clocking scheme directs themultiplexer3415 to output the configuration bits previously selected by the multiplexer3445 (i.e., the configuration bits stored incell3425c). Similarly, while themultiplexer3445 is switching between outputting the configuration bits stored incell3425cand the bits stored incell3425d, the clocking scheme directs themultiplexer3415 to output the configuration bits previously selected by the multiplexer3440 (i.e., the configuration bits stored incell3425b).

When implementing a design that has a primary clock rate of ×MHZ through a four sub-cycle implementation that effectively operates at 4×MHZ, this clocking scheme allows the configuration bits to be read from the storage cells at an effective rate of 4×MHZ without the need for a 4×MHZ clock. Some embodiments globally distribute the differential pair of CLK andCLK signals, while locally generating the differential signals STO,STO, ST1, andST1. Examples of such distribution and generation are further described in U.S. patent application entitled “Configurable IC with Interconnect Circuits that also Perform Storage Operations”, which was incorporated above.

As mentioned above,FIG. 32 conceptually illustrates themultiplexer3220 as a separate multiplexer that is after themultiplexer3210. However, some embodiments implement themultiplexer3220 in conjunction with the multiplexer structure of themultiplexer3210 that is used to generate the signal S0 and its complement.FIG. 36 illustrates a CPL implementation of this combinedmultiplexer structure3600 of some embodiments.

Thismultiplexer structure3600 outputs the select signal S0′, which is provided to the first select line of themultiplexer3205. Thecircuit3600 illustrated inFIG. 36 can be divided into five sections, which are (1)storage cell section3605, (2) first two-to-onemultiplexer stage3610, (3) a second two-to-onemultiplexer stage3620, (4) a third two-to-onemultiplexer section3625, and (5) a second pull-uptransistor stage3630.

Thestorage cell section3605 is identical to thestorage cell section3405 of thecircuit3400 ofFIG. 34, with the exception that the storage cells inFIG. 36 store configuration bits for the first select signal S0, instead of storing configuration bits for the second select signal S1. In other words, each storage cell3425 provides a configuration bit3430 and the complement of this bit3435, where each such pair of bits provides the select bit signal S0 and its complement during a particular sub-cycle.

Thesecond section3610 includes two

multiplexers

3440 and3445 that are identical to the two

multiplexers

3440 and3445 of thecircuit3400 ofFIG. 34. As incircuit3400, the

multiplexers

3440 and3445 are for outputting the configuration bits from

cells

3425a,3425b,3425c, and3425d. The outputs of thesecond section3610 can be tied to four pull-up PMOS transistors, which can quickly pull-up the outputs of the

multiplexers

3440 and3445 that are high. Some embodiments, on the other hand, might not include such pull-up PMOS transistors.

Thethird section3620 includes two two-to-

one multiplexers

3640 and3645 that select respectively between the output of themultiplexer3440 and the user signal US, and between the output of themultiplexer3445 and the user signal US, based on the output UM andUM of themultiplexer3215. As mentioned above, the output of themultiplexer3215 specifies whether the select signal S0′ should be the user signal or the select signal S0 stored in the configuration data cell for the current sub-cycle.

Thefourth section3625 is one two-to-one multiplexer3415 that is driven by a clock signal CLK, which operates at twice the frequency of the signals ST0 and ST1. The signals ST0, ST1, and CLK are illustrated inFIG. 35, as described above. Also, as mentioned above, the use of the two-to-one multiplexer3415 and the signals CLK, ST0, and ST1 and their differential complements, hides all the timing of the selection of the configuration bits from the storage cells3425 behind the two-to-one multiplexer3415.

Thefifth section3630 includes two pull-up

transistors

3665 and3670 that are used to pull-up quickly the

output nodes

3655 and3660 of thefourth section3625, and to regenerate the voltage levels at the nodes that have been degenerated by the NMOS threshold drops, when these nodes need to be at a high voltage. The

nodes

3655 and3660 provide the first select signal S0′ and its complement for the first select line of themultiplexer3205.

VII. Alternate Implementation of an HUMUX

In theHUMUX3200, the first select line of themultiplexer3205 can configurably receive either a configuration-derived select signal S0 or a user signal. Hence, the operation of theHUMUX3200 can be based either on the configuration-derived select signal S0 or on the user signal.FIG. 37 illustrates another implementation of an HUMUX, which can operate based on a configuration-derived select signal S0, a first user signal US0, or a second user signal US1. Specifically, in theHUMUX3700 of this figure, the first select line of themultiplexer3205 can configurably receive either the configuration-derived select signal S0, the first user signal US0, or the second user signal US1.

TheHUMUX3700 is identical to theHUMUX3200 except that theHUMUX3700 also includes amultiplexer3705. Thismultiplexer3705 provides theHUMUX3700 with the ability to supply the first select line of themultiplexer3205 with one of two user signals US0 and US1, in addition to the select signal S0 output from themultiplexer3210. In particular, themultiplexer3705 receives the two user signals US0 and US1 as input signals. It also receives the select signal S0 along its select line. Based on the value of the select signal S0, themultiplexer3705 routes one of the two user signals US0 and US1 to themultiplexer3220, which also receives the select signal S0 as one of its input signals. Based on the value that themultiplexer3220 receives on its select line, themultiplexer3220 then outputs as select signal S0′ either the select signal S0 or the user signal that it receives. The select signal S0′ then drives the first select line of themultiplexer3205 as described above.

FIG. 37 conceptually illustrates the

multiplexer

3705,3220, and3210 as separate multiplexers. However, some embodiments implement the two

multiplexers

3705 and3220 in conjunction with the multiplexer structure of themultiplexer3210 that is used to generate the signal S0 and its complement.

FIG. 38 illustrates a CPL implementation of this combinedmultiplexer structure3600 of some embodiments. This combined multiplexer structure is identical to the combinedmultiplexer structure3600 ofFIG. 36, except that themultiplexer structure3600 also includes 2 two-to-

one multiplexers

3805 and3810.

Themultiplexer3805 selects between the user signals US0 and US1 based on the output of themultiplexer3440, while themultiplexer3810 selects between the user signals US0 and US1 based on the output of themultiplexer3445. Each multiplexer includes two transistors (3815 and3825, or3835 and3845) that receive the user signals US0 and its complement, and two transistors (3820 and3830, or3840 and3850) that receive the user signals US1 and its complement. The transistors that receive the user signal US0 and its complement are driven by the true output of the

multiplexer

3440 or3445, while the transistors that receive the user signal US1 and its complement are driven by the complement output of the

multiplexer

3440 or3445.

The true and complement outputs of each two-to-

one multiplexer

3805 and3810 is supplied to the two-to-

one multiplexers

3640 and3645 of thethird section3620. These multiplexers then select respectively between the output of themultiplexer3440 and the user signal (US0 or US1) supplied by themultiplexer3805, and between the output of themultiplexer3445 and the user signal (US0 or US1) supplied by themultiplexer3810. The selection of these

multiplexers

3640 and3645 are based on the output UM andUM of themultiplexer3215, which specifies whether the select signal S0′ should be the user signal or the select signal S0 stored in the configuration data cell for the current sub-cycle.

Several embodiments were described above by reference to examples of sub-cycle reconfigurable circuits that operate based on four different sets of configuration data. In some of these examples, a reconfigurable circuit receives its four different configuration data sets sequentially in an order that loops from the last configuration data set to the first configuration data set. Such a sequential reconfiguration scheme is referred to as a 4 “loopered” scheme.

Other embodiments, however, might be implemented as six or eight loopered sub-cycle reconfigurable circuits. In a six or eight loopered reconfigurable circuit, a reconfigurable circuit receives six or eight configuration data sets in an order that loops from the last configuration data set to the first configuration data set. Several examples of how to make a four loopered circuit into a six and eight loopered circuit and the clocking scheme for such circuits are described in the above mentioned U.S. patent application entitled “Configurable IC with Interconnect Circuits that also Perform Storage Operations”, which is filed concurrently with the present application, with attorney docket number TBUL.P0022.

VIII. Alternative Two Tiered Structure for Retrieving Data

Several circuits described above utilize a two-tiered structure for retrieving data (e.g., configuration data, etc.) on a sub-cycle basis. Examples of such circuits are the circuits illustrated inFIGS. 34,36, and38. These circuits employ multiple storage elements3425 that store multiple sets of data for multiple sub-cycles. They also include two tiers of multiplexers, where two two-to-one multiplexers (e.g.,3440 and3445) form the first tier and one two-to-one multiplexer (e.g.,3415) forms the second tier. In some circuits, the two tiers of multiplexers have intervening circuits between them, such as AND'ing transistors or gates, etc. The second-tier multiplexer runs at the clock rate CLK, while the first-tier multiplexers runs at half that rate. From the storage elements, these multiplexers together output data at a sub-cycle rate that is twice the clock rate CLK.

Some embodiments that use this two-tiered structure, build the first tier of multiplexers into the sensing circuitry of the storage elements3425.FIG. 39 illustrates an example of such an approach. Specifically, this figure illustrates four storage elements3425a-3425dthat are arranged in two

columns

3950 and3955. Each storage element stores one logical bit of data in a complementary format. This data might be configuration data, enable data, or any other data that needs to be provided to the reconfigurable IC on a sub-cycle basis.

Each of the two complementary outputs of each storage element3425 connects to a pair of stacked NMOS transistors3920 and3925. One transistor3925 in each stacked pair of NMOS transistors is part of a first tier multiplexer structure. Specifically, in the two-tiered circuit structure3900 illustrated inFIG. 39, the first tier multiplexer structure is formed by the eight transistors3925, which receive the sub-cycle signals ST0, ST1, or the complements of these signals.

Through the sub-cycle signals ST0, ST1,ST0, andST1, the multiplexer transistors3925 selectively connect the NMOS transistors3920 to the

cross-coupled PMOS transistors

3905 and3910. One pair of

PMOS transistors

3905 and3910 exists in each column and form part of the sensing amplifier for the storage elements in that column.

Specifically, when the NMOS transistors3920 associated with one storage element3425 connect to the

PMOS transistors

3905 and3910, they form a level-converting sense amplifier. This amplifier then translates the signals stored in the storage element to the bit lines3935 or3940. The circuit3900 provides the content of the storage elements through level-converting sense amplifiers, because, in some embodiments, the storage elements are storage cells that use a reduced voltage to store their data in order to conserve power. One such example of a reduced power storage cell is provided in United States Application entitled “Method and Apparatus for Reduced Power Cell,” filed concurrently with the present application, with the attorney docket number TBUL.P0020.

The bit lines3935 and3940 connect to the next stage in the circuit that they are used. Specifically, the bit lines3935 and3940 are provided to the two-to-one multiplexer3415 in thecircuit3400, while they are provided to

transistor stages

3640 and3645 in the

circuits

3600 and3800, which, in turn, connect to themultiplexer3415. As described above, themultiplexer3415 is controlled through the clock signal CLK and its complement. Accordingly, when the clock signals CLK andCLK, and the sub-cycle signals ST0, ST1,ST0, andST1, have the timing relationship illustrated inFIG. 35, the first tier multiplexer (formed by the transistors3925) and thesecond tier multiplexer3415 operate to output data from the storage elements3425 at a rate that is twice the rate of the clock signal CLK. This outputting is analogous to how thecircuit3400 outputs the S1 select signal on the sub-cycle basis that is illustrated inFIG. 35.

By building the first multiplexer stage into the sense amplifier section of the storage elements, this circuit reduces signal path delay from the storage elements. Also, it operates with storage elements that have less power consumption. Furthermore, it reduces power consumption by using NMOS transistors3920 that are not driven by full voltage levels, and sharing the

PMOS transistors

3905 and3910 that are necessary for level conversion between two storage elements.

The two-tiered structure of the circuit3900 ofFIG. 39 can be easily extended to six and eight loopered structures. For a six loopered structure, all that needs to be done is to stack another pair of storage elements above

elements

3425cand3425d, and to drive the transistors3925 with two sets of three one-hot signals. Similarly, for an eight loopered structure, all that needs to be done is to stack two pairs of storage elements on top of

elements

3425cand3425d, and to drive the transistors3925 with the two sets of four one-hot signals. These sets of one-hot signals are further described in the U.S. patent application entitled “Configurable IC with Interconnect Circuits that also Perform Storage Operations”, which is filed concurrently with the present application, with attorney docket number TBUL.P0022.

IX. Configurable IC Architectures

Different embodiments of the invention are implemented in different configurable IC's with different architectures.FIGS. 40-45 illustrate the architecture of some embodiments of the invention. As shown inFIG. 40, this architecture is formed by numerousconfigurable tiles4005 that are arranged in an array with multiple rows and columns. InFIGS. 40-45, each configurable tile includes a sub-cycle reconfigurable three-input LUT4010, three sub-cycle reconfigurable input-

select multiplexers

4015,4020, and4025, and two sub-cycle

reconfigurable routing multiplexers

4030 and4035. Other configurable tiles can include other types of circuits, such as memory arrays instead of logic circuits.

InFIGS. 40-45, an input-select multiplexer is an interconnect circuit associated with theLUT4010 that is in the same tile as the input select multiplexer. One such input select multiplexer receives several input signals for its associated LUT and passes one of these input signals to its associated LUT.

InFIGS. 40-45, a routing multiplexer is an interconnect circuit that at a macro level connects other logic and/or interconnect circuits. In other words, unlike an input select multiplexer in these figures that only provides its output to a single logic circuit (i.e., that only has a fan out of 1), a routing multiplexer in some embodiments either provides its output to several logic and/or interconnect circuits (i.e., has a fan out greater than 1), or provides its output to other interconnect circuits.

FIGS. 41-45 illustrate the connection scheme used to connect the multiplexers of one tile with the LUT's and multiplexers of other tiles. This connection scheme is further described in U.S. Application entitled “Configurable IC with Routing Circuits with Offset Connections”, filed concurrently with this application with attorney docket number TBUL.P0036. This application is incorporated herein by reference.

In the architecture illustrated inFIGS. 40-45, each tile includes one three-input LUT, three input-select multiplexers, and two routing multiplexers. Other embodiments, however, might have a different number of LUT's in each tile, a different number of inputs for each LUT, a different number of input-select multiplexers, and/or a different number of routing multiplexers. For instance, some embodiments might employ an architecture that has in each tile: one three-input LUT, three input-select multiplexers, and eight routing multiplexers. Several such architectures are further described in the above-incorporated patent application.

In some embodiments, the examples illustrated inFIGS. 40-45 represent the actual physical architecture of a configurable IC. However, in other embodiments, the examples illustrated inFIGS. 40-45 topologically illustrate the architecture of a configurable IC (i.e., they show connections between circuits in the configurable IC, without specifying (1) a particular geometric layout for the wire segments that establish the connection, or even (2) a particular position of the circuits). In some embodiments, the position and orientation of the circuits in the actual physical architecture of a configurable IC is different than the position and orientation of the circuits in the topological architecture of the configurable IC. Accordingly, in these embodiments, the IC's physical architecture appears quite different than its topological architecture. For example,FIG. 46 provides one possible physical architecture of theconfigurable IC4000 illustrated inFIG. 40. This and other architectures are further described in the above-incorporated patent application.

X. Configurable IC and System

Some embodiments described above are implemented in configurable IC's that can compute configurable combinational digital logic functions on signals that are presented on the inputs of the configurable IC's. In some embodiments, such computations are state-less computations (i.e., do not depend on a previous state of a value).

Some embodiments described above are implemented in configurable IC's that can perform a continuous function. In these embodiments, the configurable IC can receive a continuous function at its input, and in response, provide a continuous output at one of its outputs.

FIG. 47 illustrates a portion of aconfigurable IC4700 of some embodiments of the invention. As shown in this figure, this IC has aconfigurable circuit arrangement4705 and I/O circuitry4710. Theconfigurable circuit arrangement4705 can be any of the invention's configurable circuit arrangements that were described above. The I/O circuitry4710 is responsible for routing data between theconfigurable circuits4715 of thearrangement4705 and circuits outside of the arrangement (i.e., circuits outside of the IC, or within the IC but outside of the arrangement4705). As further described below, such data includes data that needs to be processed or passed along by the configurable circuits.

The data also includes in some embodiments configuration data that configure the circuits to perform particular operations.FIG. 48 illustrates a more detailed example of this. Specifically, this figure illustrates aconfiguration data pool4805 for theconfigurable IC4700. This pool includes N configuration data sets (CDS). As shown inFIG. 48, the input/output circuitry4710 of theconfigurable IC4700 routes different configuration data sets to different configurable circuits of theIC4700. For instance,FIG. 48 illustratesconfigurable circuit4845 receiving

configuration data sets

1,3, and J through the I/O circuitry, whileconfigurable circuit4850 receivesconfiguration data sets3, K, and N−1 through the I/O circuitry. In some embodiments, the configuration data sets are stored within each configurable circuit. Also, in some embodiments, a configurable circuit can store multiple configuration data sets so that it can reconfigure quickly by changing to another configuration data set. In some embodiments, some configurable circuits store only one configuration data set, while other configurable circuits store multiple such data sets.

A configurable IC of the invention can also include circuits other than a configurable circuit arrangement and I/O circuitry. For instance,FIG. 49 illustrates a system on chip (“SoC”) implementation of aconfigurable IC4900. This IC has aconfigurable block4950, which includes aconfigurable circuit arrangement4705 and I/O circuitry4710 for this arrangement. It also includes aprocessor4915 outside of the configurable circuit arrangement, amemory4920, and abus4910, which conceptually represents all conductive paths between theprocessor4915,memory4920, and theconfigurable block4950. As shown inFIG. 49, theIC4900 couples to abus4930, which communicatively couples the IC to other circuits, such as an off-chip memory4925.Bus4930 conceptually represents all conductive paths between the components of theIC4900.

Thisprocessor4915 can read and write instructions and/or data from an on-chip memory4920 or anoffchip memory4925. Theprocessor4915 can also communicate with theconfigurable block4950 throughmemory4920 and/or4925 throughbuses4910 and/or4930. Similarly, the configurable block can retrieve data from and supply data to

memories

4920 and4925 through

buses

4910 and4930.

Instead of, or in conjunction with, the system on chip (“SoC”) implementation for a configurable IC, some embodiments might employ a system in package (“SiP”) implementation for a configurable IC.FIG. 50 illustrates onesuch SiP5000. As shown in this figure,SiP5000 includes four IC's5020,5025,5030, and5035 that are stacked on top of each other on asubstrate5005. At least one of these IC's is a configurable IC that includes a configurable block, such as theconfigurable block4950 ofFIG. 49. Other IC's might be other circuits, such as processors, memory, etc.

As shown inFIG. 50, the IC communicatively connects to the substrate5005 (e.g., through wire bondings5060). These wire bondings allow the IC's5020-5035 to communicate with each other without having to go outside of theSiP5000. In some embodiments, the IC's5020-5035 might be directly wire-bonded to each other in order to facilitate communication between these IC's. Instead of, or in conjunction with the wire bondings, some embodiments might use other mechanisms to communicatively couple the IC's5020-5035 to each other.

As further shown inFIG. 50, the SiP includes a ball grid array (“BGA”)5010 and a set ofvias5015. TheBGA5010 is a set of solder balls that allows theSiP5000 to be attached to a printed circuit board (“PCB”). Each via connects a solder ball in theBGA5010 on the bottom of thesubstrate5005, to a conductor on the top of thesubstrate5005.

The conductors on the top of thesubstrate5005 are electrically coupled to the IC's5020-5035 through the wire bondings. Accordingly, the IC's5020-5035 can send and receive signals to and from circuits outside of theSiP5000 through the wire bondings, the conductors on the top of thesubstrate5005, the set ofvias5015, and theBGA5010. Instead of a BGA, other embodiments might employ other structures (e.g., a pin grid array) to connect a SiP to circuits outside of the SiP. As shown inFIG. 50, ahousing5080 encapsulates thesubstrate5005, theBGA5010, the set ofvias5015, the IC's5020-5035, the wire bondings to form theSiP5000. This and other SiP structures are further described in U.S. patent application entitled “Programmable System In Package”, filed concurrently herewith attorney docket number TBUL.P0030.

FIG. 51 conceptually illustrates a more detailed example of acomputing system5100 that has anIC5105, which includes one of the invention's configurable circuit arrangements that were described above. Thesystem5100 can be a stand-alone computing or communication device, or it can be part of another electronic device. As shown inFIG. 51, thesystem5100 not only includes theIC5105, but also includes abus5110, asystem memory5115, a read-only memory5120, astorage device5125,input devices5130,output devices5135, andcommunication interface5140.

Thebus5110 collectively represents all system, peripheral, and chipset interconnects (including bus and non-bus interconnect structures) that communicatively connect the numerous internal devices of thesystem5100. For instance, thebus5110 communicatively connects theIC5110 with the read-only memory5120, thesystem memory5115, and thepermanent storage device5125.

From these various memory units, theIC5105 receives data for processing and configuration data for configuring the IC's configurable logic and/or interconnect circuits. When theIC5105 has a processor, the IC also retrieves from the various memory units instructions to execute. The read-only-memory (ROM)5120 stores static data and instructions that are needed by theIC5110 and other modules of thesystem5100. Thestorage device5125, on the other hand, is read-and-write memory device. This device is a non-volatile memory unit that stores instruction and/or data even when thesystem5100 is off. Like thestorage device5125, thesystem memory5115 is a read-and-write memory device. However, unlikestorage device5125, the system memory is a volatile read-and-write memory, such as a random access memory. The system memory stores some of the instructions and/or data that the IC needs at runtime.

Thebus5110 also connects to the input and

output devices

5130 and5135. The input devices enable the user to enter information into thesystem5100. Theinput devices5130 can include touch-sensitive screens, keys, buttons, keyboards, cursor-controllers, microphone, etc. Theoutput devices5135 display the output of thesystem5100.

Finally, as shown inFIG. 51,bus5110 also couplessystem5100 to other devices through acommunication interface5140. Examples of the communication interface include network adapters that connect to a network of computers, or wired or wireless transceivers for communicating with other devices. One of ordinary skill in the art would appreciate that any other system configuration may also be used in conjunction with the invention, and these system configurations might have fewer or additional components.

While the invention has been described with reference to numerous specific details, one of ordinary skill in the art will recognize that the invention can be embodied in other specific forms without departing from the spirit of the invention. Thus, one of ordinary skill in the art would understand that the invention is not to be limited by the foregoing illustrative details, but rather is to be defined by the appended claims.

Claims

1.-49. (canceled)

50. An integrated circuit (IC) comprising:

a particular arrangement of configurable logic circuits for configurably performing a set of logic functions, the particular arrangement comprising:

(i) a set of configurable logic circuits for performing a set of functions as determined by configuration data; and

(ii) a set of user interconnect circuits for performing a function based on user signals, wherein each user interconnect circuit comprises an input terminal set for receiving a set of input data, a select terminal set for receiving a set of user signals, and an output terminal set for outputting a subset of the set of input data, said outputting based on the received set of user signals.

51. The IC ofclaim 50, wherein a particular configurable logic circuit is a lookup table (LUT).

52. The IC ofclaim 50, wherein the set of user interconnect circuits is for performing a logic function.

53. The IC ofclaim 52, wherein the set of user interconnect circuits is for performing a Boolean operation.

54. The IC ofclaim 50, wherein a particular user signal is internally generated by circuits of the IC during runtime of the IC.

55. The IC ofclaim 50, wherein the user interconnect circuits are in a particular repeated pattern within circuits of the IC.

56. The IC ofclaim 50, wherein the particular arrangement comprises an interspersing of the user interconnect circuits amongst repeated patterns of circuits throughout the IC.

57. The IC ofclaim 56, wherein the particular arrangement comprises at least fifty user interconnect circuits.

58. The IC ofclaim 50 further comprising a second set of non-configurable logic circuits for performing a static set of logic functions.

59. The IC ofclaim 50 further comprising a configurable interconnect circuit for configurably supplying user signals produced by a particular set of configurable logic circuits to a particular user interconnect circuit, said supplying based on configuration data.

60. A configurable integrated circuit (IC) comprising:

a plurality of configurable logic circuits, each configurable logic circuit comprising:

a plurality of user interconnect circuits, each user interconnect circuit comprising a set of input terminals, a set of output terminals, and a set of select terminals for receiving a set of user-design signals that directs the user interconnect circuit to connect a subset of the input terminals to the set of output terminals; and

a configurable interconnect circuit for configurably outputting an output of at least one user interconnect circuit of the plurality of user interconnect circuits as determined by stored configuration data, wherein an output of a particular user interconnect circuit and the configurable interconnect circuit is an output for a particular logic function from the set of logic functions.

61. The configurable IC ofclaim 60, wherein the user-design signals determine a logic function from the set of logic functions for each user interconnect circuit to perform.

62. The configurable IC ofclaim 60, wherein the plurality of circuits is distributed within an array of circuits of the IC and said logic circuit is within the array.

63. The configurable IC ofclaim 60, wherein the plurality of circuits is distributed within a particular repeated pattern and said logic circuit is interspersed within the plurality of circuits.

64. The configurable IC ofclaim 60, wherein a particular user-design signal comprises signals generated by circuits of the IC during runtime of the IC.

65. A method for performing logic functions in an integrated circuit (IC), the method comprising:

performing a first plurality of logic functions by using a plurality of configurable logic circuits that are configured by configuration data; and

performing a second plurality of logic functions by using a plurality of user interconnect circuits, wherein a particular user interconnect circuit receives a set of input and outputs a subset of the received input based on a set of user signals.